基于分布式数据流的大数据信息安全评估平台研究  

Research on Big Data Information Security Evaluation PlatformBased on Distributed Data Flow

作　　者：潘锦锋 胡小琴 PAN Jin-feng;HU Xiao-qin(School of Software,Quanzhou University of Information Engineering,Quanzhou,362000,Fujian)

机构地区：[1]泉州信息工程学院软件学院,福建泉州362000

出　　处：《蚌埠学院学报》2021年第2期84-87,共4页Journal of Bengbu University

基　　金：福建省中青年教师教育科研项目(JAT190931)。

摘　　要：为了提升高效频繁集大数据的信息安全性,提出了基于分布式数据流的大数据信息安全评估平台。首先构建数据结构融合模型,采用闭频繁项特征融合的方法,实现对分布式数据流的频繁集挖掘;结合关联规则挖掘和结构化数据检测的方法,实现了对分布式数据流的融合和混合频繁项集的挖掘,提取了分布式数据流的关联规则属性参数集;通过模糊信息空间聚类和特征分布式融合的方法,将数据集加载到数据聚类中心,采用高效频繁集分布式数据流抗干扰融合模型,得到梯度信息分量和误差;根据数据的特征属性值来确定大数据信息记录的位置,提取分布式数据流的候选集、事务集,根据属性集分布实现了安全评估。仿真结果表明,采用该方法,大数据信息安全评估的自适应性较好,分布式数据流的信息融合度较高,提高了检测和安全性识别能力。In order to improve the information security of efficient and frequent big data aggregation,a big data information security evaluation platform based on distributed data flow was proposed in this paper.It constructed the data structure fusion model of big data information security evaluation,and adopted the method of closed frequent item feature fusion to realize the frequent set mining of distributed data stream.Combining with the method of association rule mining and structured data detection,it realized the fusion of distributed data stream and the mining of mixed frequent itemset,and extracted the attribute parameter set of association rules of distributed data stream.Through the method of fuzzy information space clustering and feature distributed fusion,the data set was loaded into the data clustering center,and the gradient information component and error of security information evaluation of frequent set distributed data stream were obtained by using efficient frequent set distributed data stream anti-interference fusion model.According to the characteristic attribute value of data,the location of big data information record was determined,the candidate set and transaction set of distributed data flow were extracted,and the security evaluation of big data information was realized according to the distribution of attribute set.The simulation results showed that the proposed method has good adaptability for big data information security assessment,and the information fusion degree of distributed data stream is high,and the detection and security identification capabilities of distributed data stream are improved.

关 键 词：分布式数据流 大数据 信息安全 频繁项集 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]